A Virtual Private Network (VPN) may be thought of as a private network constructed within a shared network infrastructure. In common terminology, these private networks are used by clients while the network infrastructure is supplied by providers.
Existing varieties of switched Layer-2 and/or Layer-1 VPNs have limitations affecting ease of implementation and use including:                clients must store and manipulate provider addresses;        clients need to be configured with all the provider addresses to which the client has a site attached;        clients need to know about connection restrictions, such as for closed-user-group (CUG) values, and need to signal these values when establishing connectivity;        clients encounter complexity in managing CUG rules; and        clients need to implement an appropriate Layer-2 and/or layer-1 signalling mechanism proper to the transport technology.        
In view of the foregoing, it would be desirable to provide a technique for providing generalized Layer-2 virtual private networks (GL2VPNs) which overcomes the above-described inadequacies and shortcomings.